PERSPECTIVES explore the current evidence for this theory OPINION and suggest that initial optimism must now be cautioned with a more objective view of The dopamine theory of addiction: the role of dopamine in addiction. 40 years of highs and lows Dopamine and the drug high Studying in vivo dopamine function in humans became possible in the mid‑1990s David J. Nutt, Anne Lingford-Hughes, David Erritzoe and Paul R. A. Stokes with the development of radiotracer imag- 11 Abstract | For several decades, addiction has come to be viewed as a disorder of ing techniques, such as C-raclopride positron emission tomography (PET) and the dopamine neurotransmitter system; however, this view has not led to new 123I-iodobenzamide (IBZM) single-photon treatments. In this Opinion article, we review the origins of the dopamine theory emission computed tomography (SPECT). of addiction and discuss the ability of addictive drugs to elicit the release of These tracers can be used to measure the dopamine in the human striatum. There is robust evidence that stimulants increase availability of striatal dopamine D2/3 recep- striatal dopamine levels and some evidence that alcohol may have such an effect, tors and changes in striatal dopamine levels in the synapse8 (BOX 1). but little evidence, if any, that cannabis and opiates increase dopamine levels. The crucial breakthrough in imaging Moreover, there is good evidence that striatal dopamine receptor availability and the human dopamine system in addiction dopamine release are diminished in individuals with stimulant or alcohol came in 1994, when it was demonstrated dependence but not in individuals with opiate, nicotine or cannabis dependence. that the combination of intravenous infusion These observations have implications for understanding reward and treatment of a central stimulant drug and SPECT or responses in various addictions. PET neurochemical imaging of dopamine D2/3 receptors could be used to indirectly measure dopamine release in the human Addiction is one of the biggest health prob- and primates. This research clearly placed striatum9,10 (BOX 1). The magnitude of this lems facing the world today. Each year, many dopamine as the central neurotransmit- increase was later shown to predict the millions of people die as a result of addiction ter in stimulant addiction5 and suggested euphoria10, or ‘high’ (REF. 11), produced by to substances such as tobacco and alcohol1, that it had roles in reward, motivation and the drug. This was interpreted as proving and currently available treatments for addic- incentive behaviour6. that the experience of pleasure (the reward- tion have limited efficacy and application. The next conceptual breakthrough came ing action) of stimulant drugs in humans Thus, there is a great need to better under- when a group of researchers in Sardinia, was mediated by striatal dopamine release, stand the brain mechanisms that are involved Italy, who pioneered the technique of brain just as in rats7. This was a powerful message in addiction so that new, better-targeted microdialysis in rats, discovered that a range that many researchers sought to develop. A interventions can be developed. of other drugs of abuse (that is, not only succession of other human studies followed, A major breakthrough in brain research stimulants) increased dopamine release in which showed that alcohol12,13, tobacco14, was made in the 1970s when the potential the nucleus accumbens, which is located ketamine15 and cannabis16 increase striatal role of dopamine in addiction was discov- in the ventral striatum7. This led to a gen- dopamine release in healthy participants and ered. This breakthrough stemmed from the eral theory of addiction in which addictive in non-dependent drug users, thereby pro- finding of Olds and Milner2 that rats would drugs release dopamine but psychoactive, viding support for the dopamine theory of willingly and repeatedly self-stimulate non-addictive drugs do not. The field devel- addiction (FIG. 1). particular areas in the brain with electric- oped rapidly from this point, with multiple The dopamine theory of addiction ity, a process that the researchers called replications of the early animal findings of rapidly became accepted by the field, and positive reinforcement. These areas were dopamine being released by ‘addictive’ drugs drugs that induce dopamine release were subsequently shown to comprise, in part, a and reported confirmations in humans using consequently considered to pose a risk set of dopamine neurons3, which explained neurochemical imaging. These findings led of abuse. An example of such a drug is why drugs that enhanced the actions of to immense investment in research to alter modafinil, which is used to treat narcolepsy. this neurotransmitter (for example, stimu- dopamine neurotransmitter function as a A 11C-raclopride dopamine imaging study lants) increased electrical self-stimulation4. route to treating addiction. Disappointingly, found that modafinil produced an increase A subsequent series of largely US-based despite four decades of intense research in dopamine release17. This finding was studies revealed that blocking dopamine effort, this theory has not led to new treat- interpreted to mean that modafinil carries a receptors with neuroleptic drugs impaired ments. In this Opinion article, we chart the potential risk for abuse, despite the increase the reinforcing effects of stimulants in rats history of the dopamine theory of addiction, in dopamine release not being associated NATURE REVIEWS | NEUROSCIENCE VOLUME 16 | MAY 2015 | 305 © 2015 Macmillan Publishers Limited. All rights reserved PERSPECTIVES Box 1 | Imaging dopamine receptors and dopamine release with striatal dopamine release in opiate dependence. For example, a study in indi- Imaging dopamine in the human striatum viduals addicted to heroin revealed that an Positron emission tomography (PET) and single-photon emission computed tomography (SPECT) intravenous dose of 50 mg heroin had no are quantitative radioactive imaging techniques that can be used to measure the availability of effect on striatal dopamine levels, despite receptors and transporters, as well as the release of neurotransmitters. In the case of the dopamine 31 system, radiotracers such as 11C-raclopride (for PET) and 123I-iodobenzamide (for SPECT) can producing a pronounced euphoric high . reliably measure the availability of the combination of dopamine D2 and D3 receptors (generally This finding was subsequently replicated in referred to as dopamine D2/3 receptors) in the human striatum, the brain area with the highest a study that further showed that expectation density of these receptors. This includes the ventral striatum (also known as the nucleus of a heroin reward (in the absence of actual accumbens), which is the region of the striatum that seems to be particularly involved in the heroin administration) was not associated acquisition of drug addiction. Most D2/3 radiotracers, and in particular the ones with agonist with dopamine release32. properties, such as 11C-(+)-4‑propyl‑3,4,4a,5,6,10b‑hexahydro‑2H‑naphtho[1,2‑b][1,4]oxazin‑9‑ol Various studies of nicotine mostly suggest (11C-PHNO), also have the ability to compete with the brain’s endogenous dopamine to bind to D2 that this compound causes a small increase 125 and D3 receptors . Increases in extracellular dopamine levels elicited by pharmacological in ventral striatal dopamine levels. For challenges, such as with methylphenidate9 or amphetamine10, or non-pharmacological 8 example, smoking cigarettes has been shown manipulations, such as stress, playing a video game or cue-induced drug craving , can be detected 11 as a decrease in the radiotracer binding owing to increased competition for the striatal D2/3 to produce a 7% reduction in C-raclopride receptors (reviewed in REF. 8). PET binding (when compared with smoking nicotine-free cigarettes)33, whereas ampheta- Extra-striatal dopamine imaging The development of radiotracers with higher affinities for D2 and D3 receptors — for example, mine produces a 10–20% reduction in the 34–42 18F-fallypride, (S)-N-((1‑ethyl‑2‑pyrrolidinyl)methyl)-5‑bromo‑2,3‑dimethoxybenzamide (11C-FLB binding of this ligand (FIG. 1). However, 457)126 and 11C-PHNO127 — has made it possible to image these receptors in brain regions outside another study (in a small cohort) found that the striatum (for example, in the frontal cortex), where the density of D2 and D3 receptors is intranasal nicotine administration had no lower. Radiotracers such as 11C-propyl-norapomorphine and in particular 11C-PHNO not only effect on ventral striatal dopamine release43. have high-affinity agonist properties but also have a higher affinity for dopamine D3 receptors This finding may be consistent with the idea over D2 receptors. This additional quality enables the quantification of D3 receptor availability that some of the effects of tobacco are due, and dopamine release in key areas of the brain in addiction, such as the globus pallidum, ventral in part, to the burning of tobacco-producing 128 tegmental area, amygdala and hypothalamus . substances that block monoamine oxidase B — the enzyme that degrades dopamine44. Studies of Δ9-tetrahydrocannabinol with an increase in ‘liking’ scores17 and prior Despite these inconsistencies and the (THC), the main psychoactive constituent clinical evidence showing that modafinil did fact that all of these drugs produced lower of cannabis, found that this compound had not cause reinforcing behaviour18. levels of dopamine